Previously, we showed fictive swimming in isolated lamprey spinal cords deviated from a uniform traveling wave. We asserted that the pattern of these deviations could be used to deduce functional properties of the intersegmental coordinating system. Using new statistical and theoretical methods, we are near a full description of the phase deviations and what they imply. In so doing, we have begun to generate statistical and theoretical tools to study changes in the spinal segments when interacting with descending and sensory inputs. By the end of year one we will begin to "put the system back together," that is, to reintroduce into the isolated spinal preparation other portions of the nervous system and sensorium. The goal is to begin to understand, in a vertebrate, how such systems are used adaptively by intact animals, and to deduce general principles of organization for such systems. Completion of the statistical and theoretical work involves: (a) simulations of the behavior of a chain of coupled oscillators with noise added. How does the noise propagate? How do the length and strength of the coordinating fibers affect the behavior? What statistical and time series methods are most appropriate for the analysis? We will then ask the following: 1. The phasic output of the reticulospinal (RS) neurons is said to be "in phase" with the motor output of the rostral segments, but this was with very few segments attached. If true, it could be disastrous since the relative phase angles of the activity of all 100 segments of the body must span 360 degrees of the cycle, and RS cells have powerful effects along the entire cord. We will use a preparation with brain stem and 50 segments to deduce: (a) What is the impact of the phasic activity from the RS nuclei upon the motor output of the spinal segments? Is the pattern more or less stable? Is it changed? (b) Using intracellular recording, what is the pattern of input from the spinal segments to the RS neurons? Is there some topographic map of the spinal segments along the nuclei? Or do all cells receive the same input? (c) Do the RS neurons have their own oscillations when activated with glutamate? What is the output from the reticular neurons to the spinal segments when a large complement of rostral and caudal segments provide input to the RS neurons? Is the RS neurons' output still phasic under these conditions? Does the output from the RS cells go to all spinal segments equally? Is it distributed across the entire cycle in all cells? (d) If phasic, how does the output from the RS nuclei interact with the coordination among the segments? This will be asked by modeling the interaction after the above data are collected. 2. If the brain and mechanosensory input both interact with the CPG, does the interaction change CPG output? (a) We will add mechanical forcing to the end of spinal segments with the brain attached to see how the output pattern is changed. This will also be done with the tail attached. Do the interactions stabilize or destabilize the vertebrate CPG? Do the brain/CPG/sensory interactions function as proposed in the cockroach to heighten responsiveness to unexpected perturbations and maintain some optimal frequency for the system? What are the functional consequences of the interactions? Finally, the overall goal is to see if we can establish principles of organization and function for motor systems that produce rhythmic movements.